PAFP Research Workshop

Noel L. Espallardo, MD, MSc, FPAFP

 Workshop Schedule
• 9:00-9:30 Welcome and Orientation
• 9:30-10:15 Introduction to Medical Research
• 10:15-10:30 Defining the Research Problem and Objectives
• 10:30-10:45 Searching for Medical Literature and Refine the Research Question
• 10:45-11:30 Workshop 1
• 11:30-11:45 The Research Proposal
• 11:45-12:00 Ethics in Health Research
• 12:00-1:00 Lunch
• 1:00-1:30 Navigating Epi-Info
• 1:30-3:00 Workshop 2
• 3:00-4:00 Group Reporting
Introduction to Medical
Research
Noel L. Espallardo, MD, MSc, FPAFP
 Definition of Clinical Research

• Systematic process to determine etiology,

pathophysiology, diagnosis, therapy and
prognosis of disease
• involves a particular person or group of
people or use of materials from humans such
their behavior or samples of their blood or
tissue

• It can involve epidemiological, behavioral, health

services and patient-oriented research like drug
trials or accuracies of diagnostic tests

• Shriver, 2010
 Purpose of Clinical Research

• The goal of clinical research is to improve

health

• Foundation of clinical decision making by

health care providers

• It can also provide evidence for policies

and decisions on health development
 Steps in Planning to Conduct Clinical Research

• Step 1 - Identify the Health Problem as a Research Idea

• Step 2 - Validate your Research Question
• Step 3 - Decide the Research Design
• Step 4 - Evaluate the Research Project with FINEST
• Step 5 - Develop General and Specific Objectives
• Step 6 – Develop the Methodology for Data Collection (Protocol)
• Step 7 – Collect and Analyze the Data
• Step 8 – Write the Final Report
 Steps in Planning to Conduct Clinical Research
• Step 1 - Identify the Health Problem as a Research Idea

• Prior knowledge and experience • Develop the health problem into a

research question
• Medical literature
• Research question should be PIO-M,
• Mentor’s recommendation but initially you can use the PIO first

• Example – Among post-term pregnant

women, what is the effectiveness of
dinoprostone gel for the induction of labor
in terms of successful vaginal delivery?
 Steps in Planning to Conduct Clinical Research

Step 2 - Validate your Research Question

• Search PubMed using PIO then limit to systematic reviews and clinical trial and the
past 5 years
• Medline search “dinoprostone” AND “induction of labor” AND “vaginal delivery”
limit to randomized controlled trial

• Purpose of validating the research question

• What has been found in previous researches and areas not adequately
addressed?
• What are the recommendations of the authors?
Validate the Question: Example
Validate the Question: Example
 Steps in Planning to Conduct Clinical Research

• Step 3 - Decide the Research Design

• The design will depend on the research

problem and the availability of funds and time
• Prevalence – cross-sectional
• Risk factor – case-control, cohort
• Effectiveness question – meta-analysis,
clinical trial, cohort
• Most students and residents select a cross-
sectional study design
• Be mindful of the time constraints associated
with the research project
Options for Research Design
 Case Report/Case Series

• A clinical case report is a written account of

the course of diagnosis, treatment and
outcomes for an individual patient
• adverse and beneficial effects
• recognition of new diseases
• unusual forms of common diseases
• presentation of rare diseases
• management of disease
 Example of Case Report
• Identify the health problem in a particular case
• A pregnant woman previously diagnosed to have spina bifida during her
childhood and subjected to corrective surgery grew up normally and is now
consulting in the first trimester of her pregnancy
• Identify the information to be collected
• Medline search “spina bifida” AND “pre-natal care” AND “clinical trials”
• Data collection
• Systematic review of articles
• Data analysis
• Based on the review develop summary recommendations for pre-natal care and
strength of recommendations
Guide for Writing a Case Report
 Cross-sectional Study Design

Subjects (random selection) Data collection/Questionnaire Results

Information at
certain point in
time
Start Future
 Example of Cross-sectional Study

• A school physician wants to determine the prevalence of unhealthy behavior among high
school students specifically on sedentary lifestyle, smoking and alcohol drinking

• Identify the subjects

• All high school students were asked to respond to a one-page questionnaire
• Identify the variables to be collected
• The questionnaire included student number or initials, age, sex, year level, physical
activity, smoking and alcohol drinking
• Data collection
• Self-administered questionnaire
 Data Analysis
Descriptive Analytical
• Numbers (Percent) • Numbers (Percent)
• Mean (Standard deviation) • Mean (Standard deviation)
• Chi-square
• Risk ratio/Odds ratio
• T-test
Guide for Writing a Cross-sectional Study
 Case-Control Study

Exposure Cases

Population

Exposure Controls

Past Start
 Example of Case-control Study

• A resident wants to determine if having multiple sexual partners is a risk factor

for the development of genital wart.
• Identify the subjects
• Cases are those who consulted the clinic and diagnosed to have genital wart
• Controls are those who consulted the clinic for other care
• Identify the variables to be collected
• Age, sex, presence or absence of genital wart, history of multiple sex partner
• Data collection
• Data collection during consultation with form containing the standard clinical
history, social history, sexual partners, physical examination and confirmatory
diagnosis
 Data Analysis
• Analysis Report
• Numbers (Percent)
• Mean (Standard deviation)
• Chi-square
• Risk ratio/Odds ratio

*Odds ratio = (50x45)/(15x10) = 15

Guide for Writing a Cross-control Study
 Cohort Study

Exposed Outcome

Population

Unexposed Outcome

Start Future
 Example of Cohort Study
• Gestational diabetes mellitus (GDM) is any degree of glucose intolerance with
onset or first recognition during pregnancy. Is GDM a risk factor for the
development of complicated pregnancy and delivery?
• Identify the subjects
• Pregnant women diagnosed to have GDM based on OGTT
• Pregnant women with normal OGTT
• Identify the variables to be collected
• Age, pregnancy details, presence of absence of GDM, complications of
pregnancy like pre-eclampsia, eclampsia, macrosomia, cesarean section, etc.
• Data collection
• Observation and follow-up of patients until delivery of baby
 Data Analysis
• Analysis Report
• Numbers (Percent)
• Mean (Standard deviation)
• Chi-square
• Risk ratio/Odds ratio
Guide for Writing a Cohort Study
 Randomized Controlled Trial

Treatment Outcome

Subjects R

Control Outcome

Start Future
 Example of Randomized Controlled Trial
• Induction of labor may be indicated in singleton post-term pregnancies to avoid CS.
A resident wants to compare the efficacy and safety of dinoprostone gel compared
with foley catheters for induction of labor.
• Identify the subjects
• Post-term pregnant women (at least 41 weeks AOG)
• After informed consent, randomized to dinoprostone gel or foley catheter
• Identify the variables to be collected
• Age, AOG, pregnancy history, pre-natal care data, labor monitoring and manner
of delivery, adverse events and complications of labor and delivery.
• Data collection
• Data collection form during labor and delivery
 Data Analysis
• Analysis Report
• Numbers (Percent)
• Mean (Standard deviation)
• T-tests/ANOVA
• Risk ratio/Odds ratio
Guide for Writing a Randomized Controlled Trial
 Options for Research Design

• Secondary study designs

• Systematic review
• Meta-analysis
• Analysis of large database
• Outcomes research
• Pharmacovigilance data
 Steps in Planning to Conduct Clinical Research

• Step 4 - Evaluate the Research Project with FINEST

• Feasibility – do you have the resources i.e. financial and technical capacity to do
the research
• Interesting – is this interesting to the researcher and target readers
• Novel – not a duplication of an existing research in the same population
• Ethical – respect for patient right to autonomy and privacy
• Significant – does it address a significant health problem
• Time bound – can be finished in one year
 Steps in Planning to Conduct Clinical Research
• Step 5 - Develop General and Specific Objectives

• The general objective states the general purpose or goal of the study i.e. a general
statement of what the results of the study can contribute to health care.
• Develop the specific objectives using SMART criteria
• Specific
• Measurable “To determine the rate of successful vaginal
• Attainable delivery among patients given dinoprostone
• Relevant plus expectant management compared to
• Time bound expectant management alone within 24
hours of labor induction.”
 Steps in Planning to Conduct Clinical Research

Step 6 – Develop the Methodology for Data Collection

• Parts of methodology section

• Research design
• Setting, definition of subjects, inclusion and exclusion, selection and
recruitment
• Data collection, intervention, measurements
• Sample size computation, data analysis
• Ethical issues
 Defining the Population

• Target population • Define the study population

• Study population • What or who should be the study population
• Inclusion criteria
• Exclusion criteria – those who fit the
inclusion but should be excluded because
the data they might contribute will not be
accurate

• When and where should the study population

be recruited
• How should the study population be selected
 Inclusion Criteria
• “Patients consulting for hypertension at the Family Medicine Clinic of the Philippine
General Hospital between the period January 1 to December 31, 1998 will be
recruited for the study.

• To be included they should fulfill the following criteria:

• Age 25-55 years old
• Systolic blood pressure equal or more than 140 mmHg
• Diastolic blood pressure equal or more than 90 mmHg
• No history of myocardial infarction or stroke”

• Exclusion criteria
• Those with psychological or mental health problems
• With history of allergy or any adverse reaction to the study drugs
 Sampling

• Purposive sampling include patient whenever they are available

• Random sampling is a process where everyone has the same chance of being
included.
• Existing list of the potential participants in the study
• Divide the list by the sample size needed and use the product as the
interval (systematic random sampling)
• If you need 50 subject from a list of 300 (300/50=6), you choose every 6
person in the list
MS Excel Data Analysis
 Data Collection Form
• The data collected must be relevant
to the objectives
• Data entry program or database
program should be designed at
the start of the project
• Spread sheet can also be used for
data entry
 Tips for Self-administered Questionnaire

• Data collection form should have introduction and purpose, demographic profile, main
study variables and closing statement.
• Questions must be clear, and answers are mutually exclusive and very few open-
ended questions
• Answerable between 10-20 minutes.
• Pre-test the questionnaire to around 10-20 persons

• Data collector should be trained and familiar with the data collection tools
• Questions and clarification from the subjects or respondents must be addressed
• Checked for completeness and errors corrected prior to data analysis
 Sample Size

• StatCalc in Epi-Info

• Need information on effect size, the level of statistical significance, and power
• Effect size refers to the size of the association of variables or effectiveness of
treatment i.e., estimated risk ratios or proportions/percentages
• Level of statistical significance or alpha is the probability of making a Type I
error is usually set at 0.05
• Beta is the probability of making a Type II error usually set at 0.2, and the
value 1-beta (1-0.2=0.8 or 80%) is the power of the study to correctly
rejecting the null hypothesis.
Sample Size Computation
 Data Analysis
• Descriptive statistics
• percentages, proportions, mean, median, and mode, range, variance, and
standard deviation

• Inferential statistics
• Chi-Square test used to assess relationship between two categorical variables
• Independent t-test is a parametric test that may be used to compare the means
of two groups
• Pearson’s correlation is a measure of the strength and direction of association
that exists between two continuous variables
 Ethical Issues
• Main elements for the patients
• Respect for autonomy – participation is voluntary, no element of coercion, signed
informed consent
• Full information about the study processes

• Main elements for the investigators are:

• Technically capable of doing the specific research
• Disclosure of potential conflict of interest
• Integrity in data collection, analysis and confidentiality of information
• Disclosure of study findings and adverse events
 Steps in Planning to Conduct Clinical Research
Step 7 – Collect and Analyze the Data

• Check for completeness of all the collected forms and label all forms
• Enter the data in a database or spreadsheet and check for data entry errors

• Develop dummy tables in a way that it will answer your hypothesis and objectives
• Dummy tables should address both descriptive and analytical components
• Table – Demographic and Clinical Profile of Subjects
• Table – Main Results
• Analysis with a free software like Epi-Info or RevMan for meta-analysis
• You can also use proprietary software like SPSS, STATA or SAS
 Steps in Planning to Conduct Clinical Research
Step 8 – Write the Final Report
 Steps in Planning to Conduct Clinical Research
Step 8 – Write the Final Report
Thank You
 Defining the Research
Problem and Objectives

Noel L. Espallardo, MD, MSc, FPAFP

 Research Idea
• Choosing and defining a research topic is the first
step in applying the scientific method to a
medical research problem.
• The scientific method itself can be viewed as
nothing more than organized curiosity, questions
and answers.
• standardized and efficient technique for
describing these organized steps in a way that
can be verified by other observers
 Sources of Research Idea
• Experience - day-to-day experience constantly provides opportunities to ask,
"Why are things done this way?" or "What would happen if things can be done in
another way?"
• skeptical attitude
• Local morbidity and mortality statistics - relevant in your setting
• Scientific gatherings, discussion with friends or colleagues or experts
• Medical literature
• previous published research
• A research idea can pop out of the blue anywhere or anytime and then disappear
• write it down in written or digital format that can be retrieved when needed
 Focusing the Research Idea
• Research ideas do not • Levels of research
materialize because they are • Level 1 - describing what occurs in a
vague or too broad
particular situation and the research design
• Focusing on just one research tends to be exploratory or descriptive.
problem is enough.
• Level 2 - description of the relationship or
• A research problem is a association between or among variables
specific issue, difficulty, and the method is often a descriptive
contradiction, or gap in
knowledge survey using structured questionnaires or
physiologic measurements.
• The problem statement can be
in the form of a hypothesis or • Level 3 - determine the effect of one
a research question. variable (independent) on another variable
(dependent) and method is interventional
study.
 Finalizing the Research Question

Elements of Good Research Question Tips in Finalizing the Research Question

• Biologic or theoretical rationale. • Start with research problem
• Specific population to be studied. • Formulate the initial research question
• Specific test intervention and the includes it should contain the P-
comparator population to be studied, I/E -
• Specific primary outcome of interest. intervention or exposure and O -
outcome to be measured.
• Suggestion of a study design.
• Review existing literature
• Refine the key variables
• Consider research design
 Sample Research Question
• “Among patients consulting at the local health center and diagnosed to have PTB
randomized to a clinical trial, will compliance to TB treatment be improved by
encouraging a family member to be a surrogate observer to drug intake
compared with usual clinic advise?”
Biologic rationale family members can encourage drug intake thereby
improving compliance
Specific population patients diagnosed to have PTB in a community health
center
Specific test intervention use of surrogate observer for drug intake
Specific primary outcome of interest improving compliance to TB chemotherapy
Specific comparative intervention usual clinic advise
Suggestion of a study design randomized clinical trial
 Criteria for a Good Research Question:
FINEST
• Feasibility - affected by several factors, but the most prominent may be the cost
of the project and the availability of study population.
• Interesting - view the interest first from the investigator, second from the patient
and third from the health care provider.
• Novel – contributes new information
• Ethical - not place the participants in unnecessary risk or deny them of the
necessary benefits.
• Significance - potential implications of the study results to patients, the medical
community, or society
• Time Bound - specify the expected amount of time necessary to complete the
project.
 Research Goals and Objectives
• The goal or general objective states the
general purpose of the study i.e., a general
statement of what the results of the study
can contribute to health care.
• The general objective is usually expressed
in a broad health care issue or statement.
• “The general objective of this study is
to improve the care and treatment of
hypertensive patients in family
practice.”
 Research Goals and Objectives: SMART

• Specificity - focused on a particular aspect of the

research problem
• Measurability - formulated in a way that allows for
measurement or observation Achievability -
realistic and attainable
• Relevance - contributes to addressing the research
problem
• Time-bound - consider the overall timeline of the
research project
Thank You
Searching for Medical Literature
and Refining the Clinical
Question
 Question for Clinical Assessment
• Clinical history and physical examination should be focused to getting
information on factors that might increase the probability of disease
• Review questions about “risk or harm” will lead to a more focused clinical
assessment
• POEM (population, outcome, exposure and methodology)
o Population: populations or subgroup of patients are we interested in.
o Outcome: development of the disease.
o Exposure: Risk factor in history or physical examination finding
o Method: risk factor is best answered using a cohort or case-control study.
Cross-sectional studies do not establish a cause-and-effect relationship of risk
factors.
 Question for Diagnostic Test
• A diagnostic test (laboratory, pathologic test or imaging) is a means of
determining whether a patient has the disease
• Questions about diagnosis are concerned with the accuracy or performance of a
diagnostic test or test strategy
• The PDCO framework
o Population: populations or subgroup of patients
o Diagnostic test: the test or test strategy being evaluated.
o Comparator: usually the reference standard
o Outcome: usually sensitivity, specificity, predictive values, likelihood ratios, or
a receiver operating characteristic (ROC) curve.
 Question on Pharmacologic Intervention

• Structure for developing questions about interventions is the PICO (population,

intervention, comparator and outcome)
o Population: populations or subgroup of patients.
o Intervention: intervention, (pharmacologic or non-pharmacologic approach
o Comparators: main alternative(s) to compare with the intervention being
considered
o Outcome: outcomes to be considered
 PICO Variants
• PICOT
• P - population, I – intervention, C – comparator, O – outcome and T – time
frame

• “Among adult patients with DM type 2, what is the effectiveness of metformin

vs. gliburide in controlling HA1c in three months?”

• PICOS
• P - population, I – intervention, C – comparator, O – outcome and S –
setting

• “Among adult patients with DM type 2, what is the effectiveness of metformin

vs. gliburide in controlling HA1c in family practice?”
 SPICE: Public Health Programs
• S – Setting: City-wide health system
• P – Perspective: Public fund
• I – Intervention: Early GDM screening
• C – Comparison: Standard GDM screening
• E – Evaluation: Health economic analysis
Booth, 2004

• In a city-wide health system, what is the cost-effectiveness of early GDM

screening vs. standard GDM screening from the perspective of public funding.
 ECLIPSE: Social Health Services
• E – Expectation: Why does the user want the information?
• C - Client Group: For whom is the service intended?
• L – Location: Where is the service physically sited?
• I – Impact: What would represent success? This component is similar to
outcomes of the PICO framework.
• P – Professionals: Who provides or improves the service?
• SE – Service: What type of service is under consideration?

“In a family with a child diagnosed to have ALL, what is the effectiveness of social
support provided by patient groups or foundations in the community that can be
used for advocacy.”

Wildridge and Bell, 2002

 Literature Search

• Search must be structured and logical, and each step must be recorded
(Rodseth and Marais, 2016)

• Based on the key terms from the clinical questions (PICO, POEM, ECLIPSE,
SPICE)
• free-text terms and MeSH
• Asterisk (*) marks for example, if you use something like "diabet*”
(Tawfik, 2019) (Tawfik, 2019)
 Literature Search
• Search must be structured and logical, and each step must be recorded
(Rodseth and Marais, 2016)

• Boolean Logic connectors "AND", "OR", and "NOT"

• Trial search by stepwise combination start with “P AND I”, then “P AND I
AND O”
(Basu, 2017)

• Some databases do not support Boolean or quotation

Stepwise Search
Stepwise Search
Stepwise Search
 Frame a Question: Validate the Question
• Simple search in PubMed or Google
Scholar with general search terms
• Preliminary search to determine
if articles are available
• Gain familiarity topic
Tawfik, 2019

• Refine the question to make sure it is

now correctly focused
• Find if there is a recent meta-
analysis on the topic
• Differentiate your focus
Rodseth and Marais, 2016
Other Databases to be Search
 Retrieving Full Text Articles
• Many search engines provide links for free to access full text articles.
• ResearchGate, Google Scholar, Google may have full text or purchase
Tawfik, 2019

• Searching reference lists of included articles, citation tracking, follow all

“related to” or “similar” articles
Tawfik, 2019
• Second level screening
• Brief review of the methods and results in the full text
• Careful record must be kept of the number of excluded abstracts and
trials as well as the reasons for their exclusion
• Selection process, must be included in the report (PRISMA Diagram)
Rodseth and Marais, 2016
 Retrieving the Full Text
• To assess the quality and extract data
from the evidence the full text must
be retrieved
• PubMed free full text
• Some websites
• Free Medical Journals
(LinksMedicus),
• Genamics JournalSeek
• Sci-Hub.
• You can also ask assistance from an
institution for its library service.
Thank You
Workshop 1
 Workshop 1
• Form 4 groups
• Develop a research problem on 4 case scenarios
• Conduct a literature search to refine the research problem
• Finalize the research question
• Select the study design to answer the research question
 Case Scenarios
Case Scenario #1
• In your hospital, you notice that there is a significant number of absentees from
the residents lately. This happens even though you have enough number of
residents to go on 24-hr duty every 3 days and the patient load at the OPD (30
per day) and ward (10 admitted) seem to be manageable. The common excuse is
too much stress and burnout from work duties. What research project can you
develop to address this problem?

Case Scenario #2
• While examining a 40 y/o male patient diagnosed to have diabetes mellitus, you
notice that despite compliance to medication his FBS is still uncontrolled. When
you check on his diet, the patient still seems to have excess calorie intake. He said
he cannot avoid the urge to taste delicious food served at home and work
cafeteria. What research project can you develop to address this problem?
 Case Scenarios
Case Scenario #3
• In one of your duties at the emergency room, you attended to a 16 y/o female
asthmatic patient in moderate acute exacerbation. The patient was relieved after
2 hours with bronchodilator nebulization and single dose steroid injection. On
further history, you learned that the patient is taking inhaled bronchodilator
whenever she feels dyspneic. What research project can you develop to address
this problem?

Case Scenario #4
• During your community rotation, you noticed that the overall immunization rate
for polio, measles and DPT in the community was only 70% based on the report
of the barangay midwife. This happens despite the availability of vaccines. One of
the reasons presented was the fear of side effects. What research project can you
develop to address this problem?
Research Proposal
Noel L. Espallardo, MD, MSc, FPAFP
 Developing a Research Proposal

• Research proposal must be detailed and

comprehensive and should be completed long before
the study begins
• Research protocol serves three main purposes
• will help clarify the goals of the study and methods
of investigation.
• provides an operational guide for the entire
research team.
• to obtain approval to proceed with the study from
either funding source, institutional review board, or
institution/department before a study may begin.
 Developing a Research Proposal

• General Plan • Emphasize

• What is the primary goal of the study? • Standard and innovative procedures
• What makes this study important? • advantages and disadvantages of
• What has previous research shown in methods to be used
this area? • limitations of the methods
• How is the study to be carried out? • anticipated difficulties
• This last question has many • alternative approaches in case of
elements that must be clearly experimental obstacles
stated. • sequence of experimental events
• written in language comprehensible (make a flow chart if necessary)
to all team members • procedures used to analyze the data
 Content of Research Proposal
• Title page - includes the title of the project, names of investigator/co-
investigator(s) and their institutional affiliation.
• Introduction - include general discussion of the topic, the previous studies and
how this proposed research will complement the previous ones, and the
significance of the proposed research.
• Introduce the research problem and provide background information to
contextualize the study.
• Summarize relevant studies and research that have been conducted in the
same or related areas and identify gaps, controversies, or limitations
• Highlight the innovation or new information this research will contribute
• Research question or hypothesis
 Content of Research Proposal
• Research objectives - overall goals and specific objectives of your research
• Methodology section - specific procedures to be carried out on human subjects in
sufficient detail to permit evaluation of the risks and replication by other
researchers.
• Research design - overall approach and design of your research (e.g.,
experimental, observational, survey, case study) and why the chosen design is
appropriate for addressing your research question.
• Participants - the target population or sample for your study
• Data collection - methods you'll use to collect data
• Data analysis - outline the statistical or analytical methods you'll employ to
analyze the collected data.
 Content of Research Proposal
• Study population - how will they be recruited, inclusion and exclusion criteria etc.
• The setting or facility where the subjects will be gathered
• how random selection or randomization will be done
• Procedures to be performed
• Intervention if clinical trial
• Exposures and outcomes and how they will be measured like laboratory tests
• If there are questionnaires or data collection instruments, describe the
instrument, how it was validated, or pilot tested.
• Data analysis – proposed presentation and analysis of data
• Ethical consideration - precautions to minimize risks
• informed consent from participants
 Content of Research Proposal
• Expected results
• Significance and contribution to current knowledge
• Project timeline - Gantt chart
• Budget - estimated costs associated with your research. Include expenses related
to materials, equipment, participant compensation, etc.
• Summary
• Funding and other support
• References
Thank You
Ethics in Health Research
Noel L. Espallardo, MD, MSc, FPAFP
 Ethics in Health Research
• Critical aspect to ensure the well-being, rights, and dignity of research
participants while promoting the generation of valuable and reliable scientific
knowledge.
• Research objective does not overshadow society’s health care goals and
protection of individual patient’s rights.
• To guarantee that protection
• Researchers should be familiar with and follow the institution’s ethical
guidelines and policies in conducting research
• Review boards are necessary to ensure that proposed studies do not violate
patient rights.
 Institutional and Ethical Review
Committees
• Institutional Review Committee/ Institutional Review Board (IRC/IRB) or Ethical
Review Committee/ Ethical Review Board (ERC/ERB).
• Group of persons formally designated by an institution to review medical
research involving human subjects
• Detailed procedures vary from one institution to another because of
difference in the capacity of the facility, setting and other factors
• Most ERCs approve a research protocol if
• the risks to the subjects are outweighed by the benefits to the subjects,
• the knowledge to be gained warrant a decision to accept the risks,
• free and informed consent will be obtained
• the rights and welfare of any such subjects will be protected
 Summary of Ethical Considerations
▪ Beneficence and maleficence - Research should aim to maximize benefits and
minimize harm to participants.
▪ Justice – Researchers must ensure the fair distribution of the benefits and
burdens of research. Special attention should be given to vulnerable populations
to prevent exploitation and ensure equitable access to research opportunities.
▪ Informed consent - Researchers must obtain voluntary and informed consent
from participants before involving them in the study.
▪ Respect for autonomy - Participants have the right to make independent
decisions about their involvement in research. Researchers should respect the
autonomy of participants and allow them to withdraw from the study at any time
without facing negative consequences.
 Summary of Ethical Considerations
▪ Data privacy and confidentiality - Data should be anonymized, and researchers
should take measures to prevent unauthorized access to sensitive information.
▪ Scientific integrity - Researchers should conduct their studies with honesty and
integrity, avoiding fabrication, falsification, and plagiarism.
▪ Community engagement - In community-based research, engaging with the
community is essential to understand its needs, values, and concerns.
▪ Institutional review board approval and monitoring - ethical review by an IRB or
ERB is often required before initiating research to ensure that it meets ethical
standards.
▪ The IRB/ERB evaluates the study design, informed consent process, and other
ethical considerations.
 Content of Informed Consent
• A clear and concise statement of the purpose of the research study and why the
individual is being asked to voluntarily participate in the study
• The study and procedures to be performed should be described, like treatment
and laboratory procedures and reasonable expectations of time commitment
• Benefits, risks and discomforts of each procedure should be included, as well as
degree of risk
• Alternatives, including non-participation or possibility of withdrawal without
negative consequence
• Management after completion of the study in studies that involve therapeutic
interventions.
• Indicate compensation to subjects.
Thank You
Navigating Epi Info
 Epi-Info
• Epi Info is a software designed for health
practitioners and researchers
• It provides for easy data entry form and
database construction, and data analyses
• Can be used for outbreak investigation,
developing small to mid-sized disease
surveillance
• Education in epidemiology and public health

• [Link]
 Define the Problem: Preventing Postpartum
Hemorrhage
• Philippines is still among the countries with
high maternal deaths
• 114 deaths per 100,000 live births were
registered in the country
• target is 52 deaths per 100,000 live
births
• 30% were reportedly associated with PPH

• Among patients at high risk for post-partum

hemorrhage undergoing normal vaginal
delivery, will carboprost given at third stage
of labor prevent postpartum hemorrhage?
 Sample Size Computation Using Epi-Info
• Incidence of post-partum
hemorrhage
• 35% PPH in oxytocin
• 11% PPH in carboprost

• Mean Blood Loss

• 610 ml in oxytocin
• 438 ml in carboprost
Statistical Analysis Using Epi-Info
Sample Size Computation Using Epi-Info
Sample Size Computation Using Epi-Info
Sample Size Computation Using Epi-Info
 Data Collection
• Most time consuming and expensive
step
• Prioritize efficiency in data collection
and management
• Address missing data

• The data collected must be relevant

to the objectives
• Data entry program should be
designed at the start of the
project
• Spread sheet or database program
can also be used for data entry
 Steps in Data Analysis
• Step 1 – Define the Problem
• Revisit the research question to be answered.
• Revisit the objectives
• Step 2 – Develop hypothetical/dummy tables and figures that will answer the
objectives
• Table or figure for descriptive profile of subjects included
• Table or figure for the analysis of main outcomes
• Table or figure for the analysis of secondary outcomes or secondary analysis
• Step 3 – Analyze the data using Epi-info
• Step 4 – Fill-in the tables
• Step 5 – Write the narrative for each tables and figures to make complete the
results section of the paper
 Define the Problem: Preventing Postpartum
Hemorrhage
• Among patients at high risk for post-partum
hemorrhage undergoing normal vaginal
delivery, will carboprost given at third stage
of labor prevent postpartum hemorrhage?

• Objective
• Determine the effectiveness of carboprost in
terms of preventing post-partum
hemorrhage and anemia and decreasing
blood loss
 Start with the Hypothetical Tables
Table 1. Baseline Table 2. Blood Loss and Post-partum Hgb
Carboprost Placebo p Value Carboprost Placebo p Value
Mean Age Blood Loss
Gravidity Post-partum Hgb
One
Two
Three
Four or more
Parity Table 3. Prevention of PPH and PP Anemia
One Carboprost Placebo RR (95% CI)
Two
Post-partum
Three
Hemorrhage
Four or more
Post-partum
Baseline Hgb
Anemia
 Epi-Info Data Entry and Management
• This step-by-step procedure can be simplified and made a lot easier with statistical
software like Epi-Info.
• Step 1. Enter the data into a spreadsheet.
• In this case, you simply enter the data from the data collection form.
• First row should be data labels
• 2nd row and succeeding rows are the observation from individual subjects
• Step 2. Read the file in Epi-Info
• Step 3. Select and run the desired statistical analysis
• Descriptive statistics
• Inferential statistics
• Step 4. Interpret the results.
Data Layout in MS Excel
 Statistical Analysis Using Epi-Info
• RCT on the effectiveness of
carboprost in the management of
third stage labor to prevent post-
partum hemorrhage

• Objective
1. Determine the effectiveness of
carboprost in terms of preventing
post-partum hemorrhage and
anemia and decreasing blood loss
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
 Descriptive Statistics

• Descriptive statistics methods for organizing a large set of numbers to a

few, informative numbers that will describe the data.
• Descriptive statistics use measures of central tendency (mean) and
dispersion (standard deviation)
• If nominal, frequency distribution can be represented as a table, or
a figure called histogram.
• If in percent, the figure of the distribution can be represented as a
pie chart.
 Start with the Hypothetical Tables
Table 1. Baseline Table 2. Blood Loss and Post-partum Hgb
Carboprost Oxytocin p Value Carboprost Oxytocin p Value
Mean Age Blood Loss
Gravidity Post-partum Hgb
One
Two
Three
Four or more
Parity Table 3. Prevention of PPH and PP Anemia
One Carboprost Oxytocin RR (95% CI)
Two
Post-partum
Three
Hemorrhage
Four or more
Post-partum
Baseline Hgb
Anemia
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
 Start with the Hypothetical Tables
Table 1. Baseline Table 2. Blood Loss and Post-partum Hgb
Carboprost Oxytocin p Value Carboprost Oxytocin p Value
Mean Age 28.13 (4.2) 28.7 (4.4) Blood Loss
Gravidity Post-partum Hgb
One
Two
Three
Four or more
Parity Table 3. Prevention of PPH and PP Anemia
One Carboprost Oxytocin RR (95% CI)
Two
Post-partum
Three
Hemorrhage
Four or more
Post-partum
Baseline Hgb
Anemia
 Descriptive Statistics: Frequency Distribution

▪ In frequency distribution, the frequency of occurrence is given for each value of the
variable.
Carboprost Oxytocin p Value
Mean Age 28.13 (4.2) 28.7 (4.4)
Gravidity
One
Two
Three
Four or more
Parity
One
Two
Three
Four or more
Baseline Hgb
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
 Start with the Hypothetical Tables
Table 1. Baseline Table 2. Blood Loss and Post-partum Hgb
Carboprost Oxytocin p Value Carboprost Oxytocin p Value
Mean Age 28.13 (4.2) 28.7 (4.4) Blood Loss
Gravidity Post-partum Hgb
One 13 (21.6%) 11 (18.3%)
Two 19 (31.6%) 16 (26.6%)
Three 28 (46.6%) 33 (55.0%)
Four or more 0 0
Parity Table 3. Prevention of PPH and PP Anemia
Zero 13 (21.6%) 11 (18.3%) Carboprost Oxytocin RR (95% CI)
One 19 (31.6%) 16 (26.6%)
Post-partum
Two 28 (46.6%) 33 (55.0%)
Hemorrhage
Three 0 0
Four or more 0 0 Post-partum
Anemia
Baseline Hgb
 Inferential Statistics: Hypothesis Testing
Table 1. Baseline Table 2. Blood Loss and Post-partum Hgb
Carboprost Oxytocin p Value Carboprost Oxytocin p Value
Mean Age 28.13 (4.2) 28.7 (4.4) Blood Loss
Post-partum Hgb
Gravidity
One 13 (21.6%) 11 (18.3%)
Two 19 (31.6%) 16 (26.6%)
Three 28 (46.6%) 33 (55.0%)
Four or more 0 0
Table 3. Prevention of PPH and PP Anemia
Parity
Zero 13 (21.6%) 11 (18.3%) Carboprost Oxytocin RR (95% CI)
One 19 (31.6%) 16 (26.6%) Post-partum
Two 28 (46.6%) 33 (55.0%) Hemorrhage
Three 0 0
Four or more 0 0 Post-partum
Anemia
Baseline Hgb
 Analysis of Continuous Variables: Two Groups
▪ Unpaired Samples of Continuous Data
• The unpaired t test compares the means of two groups and determines the probability that the
observed difference did not occur by chance.
• A small p value (less than 0.05) suggests that such a difference is unlikely to occur by chance.
• We would say that the sample means are significantly different.
• If the data is not normally distributed or not equal variance, you should use the Mann-Whitney
Rank Sum test.

▪ Paired Samples of Continuous Data

• Paired t-test is used when two measurements from the same group (before and after) is being
compared.
• A small p value (less than 0.05) suggests that such a difference is unlikely to occur by chance.
• We would say that the sample mean is significantly different from each other.
• If the data fails the normality or not equal variance, use the Wilcoxon Signed Rank test.
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
 Start with the Hypothetical Tables
Table 1. Baseline Table 2. Blood Loss and Post-partum Hgb
Carboprost Oxytocin p Value Carboprost Oxytocin p Value
Mean Age 28.13 (4.2) 28.7 (4.4) 0.42* Blood Loss
Gravidity Post-partum Hgb
One 13 (21.6%) 11 (18.3%)
Two 19 (31.6%) 16 (26.6%)
Three 28 (46.6%) 33 (55.0%)
Four or more 0 0
Parity Table 3. Prevention of PPH and PP Anemia
Zero 13 (21.6%) 11 (18.3%) Carboprost Oxytocin RR (95% CI)
One 19 (31.6%) 16 (26.6%)
Post-partum
Two 28 (46.6%) 33 (55.0%)
Hemorrhage
Three 0 0
Four or more 0 0 Post-partum
Anemia
Baseline Hgb
* t-test
 Start with the Hypothetical Tables
Table 1. Baseline Table 2. Blood Loss and Post-partum Hgb
Carboprost Oxytocin p Value Carboprost Oxytocin p Value
Mean Age 28.13 (4.2) 28.7 (4.4) 0.42* Blood Loss
Gravidity Post-partum Hgb
One 13 (21.6%) 11 (18.3%)
Two 19 (31.6%) 16 (26.6%)
Three 28 (46.6%) 33 (55.0%)
Four or more 0 0
Parity Table 3. Prevention of PPH and PP Anemia
Zero 13 (21.6%) 11 (18.3%) Carboprost Oxytocin RR (95% CI)
One 19 (31.6%) 16 (26.6%)
Post-partum
Two 28 (46.6%) 33 (55.0%)
Hemorrhage
Three 0 0
Four or more 0 0 Post-partum
Anemia
Baseline Hgb
* t-test
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
 Start with the Hypothetical Tables
Table 1. Baseline Table 2. Blood Loss and Post-partum Hgb
Carboprost Oxytocin p Value Carboprost Oxytocin p Value
Mean Age 28.13 (4.2) 28.7 (4.4) 0.42* Blood Loss
Gravidity Post-partum Hgb
One 13 (21.6%) 11 (18.3%) 0.65**
Two 19 (31.6%) 16 (26.6%)
Three 28 (46.6%) 33 (55.0%)
Four or more 0 0
Parity Table 3. Prevention of PPH and PP Anemia
Zero 13 (21.6%) 11 (18.3%) 0.56** Carboprost Oxytocin RR (95% CI)
One 19 (31.6%) 16 (26.6%)
Post-partum
Two 28 (46.6%) 33 (55.0%)
Hemorrhage
Three 0 0
Four or more 0 0 Post-partum
Anemia
Baseline Hgb
* t-test, **chi-square
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
 Start with the Hypothetical Tables
Table 1. Baseline Table 2. Blood Loss and Post-partum Hgb
Carboprost Oxytocin p Value Carboprost Oxytocin p Value
Mean Age 28.13 (4.2) 28.7 (4.4) 0.42* Blood Loss
Gravidity Post-partum Hgb
One 13 (21.6%) 11 (18.3%) 0.65**
Two 19 (31.6%) 16 (26.6%)
Three 28 (46.6%) 33 (55.0%)
Four or more 0 0
Parity Table 3. Prevention of PPH and PP Anemia
Zero 13 (21.6%) 11 (18.3%) 0.56** Carboprost Oxytocin RR (95% CI)
One 19 (31.6%) 16 (26.6%)
Post-partum
Two 28 (46.6%) 33 (55.0%)
Hemorrhage
Three 0 0
Four or more 0 0 Post-partum
Anemia
Baseline Hgb 120.4 (6.1) 119.5 (5.6) 0.40*
* t-test, **chi-square
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
 Start with the Hypothetical Tables
Table 1. Baseline Table 2. Blood Loss and Post-partum Hgb
Carboprost Oxytocin p Value Carboprost Oxytocin p Value
Mean Age 28.13 (4.2) 28.7 (4.4) 0.42 Blood Loss 350 (116) 486 (97) <0.01*
Gravidity Post-partum Hgb
One 13 (21.6%) 11 (18.3%) 0.65 * t-test
Two 19 (31.6%) 16 (26.6%)
Three 28 (46.6%) 33 (55.0%)
Four or more 0 0
Parity
Table 3. Prevention of PPH and PP Anemia
Zero 13 (21.6%) 11 (18.3%) 0.56 Carboprost Oxytocin RR (95% CI)
One 19 (31.6%) 16 (26.6%)
Post-partum
Two 28 (46.6%) 33 (55.0%)
Hemorrhage
Three 0 0
Four or more 0 0 Post-partum
Anemia
Baseline Hgb 120.4 (6.1) 119.5 (5.6) 0.40
* t-test, **chi-square
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
 Start with the Hypothetical Tables
Table 1. Baseline Table 2. Blood Loss and Post-partum Hgb
Carboprost Oxytocin p Value Carboprost Oxytocin p Value
Mean Age 28.13 (4.2) 28.7 (4.4) 0.42 Blood Loss 350 (116) 486 (97) <0.01*
Gravidity Post-partum Hgb 119 (6) 112 (11) <0.01*
One 13 (21.6%) 11 (18.3%) 0.65 * t-test
Two 19 (31.6%) 16 (26.6%)
Three 28 (46.6%) 33 (55.0%)
Four or more 0 0
Parity
Table 3. Prevention of PPH and PP Anemia
Zero 13 (21.6%) 11 (18.3%) 0.56 Carboprost Oxytocin RR (95% CI)
One 19 (31.6%) 16 (26.6%)
Post-partum
Two 28 (46.6%) 33 (55.0%)
Hemorrhage
Three 0 0
Four or more 0 0 Post-partum
Anemia
Baseline Hgb 120.4 (6.1) 119.5 (5.6) 0.40
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
 Start with the Hypothetical Tables
Table 1. Baseline Table 2. Blood Loss and Post-partum Hgb
Carboprost Oxytocin p Value Carboprost Oxytocin p Value
Mean Age 28.13 (4.2) 28.7 (4.4) 0.42 Blood Loss 350 (116) 486 (97) <0.01*
Gravidity Post-partum Hgb 119 (6) 112 (11) <0.01*
One 13 (21.6%) 11 (18.3%) 0.65 * t-test
Two 19 (31.6%) 16 (26.6%)
Three 28 (46.6%) 33 (55.0%)
Four or more 0 0
Parity
Table 3. Prevention of PPH and PP Anemia
Zero 13 (21.6%) 11 (18.3%) 0.56 Carboprost Oxytocin RR (95% CI)
One 19 (31.6%) 16 (26.6%)
Post-partum 5 (8.3%) 24 (40.0%) 0.28 (0.12-
Two 28 (46.6%) 33 (55.0%)
Hemorrhage 0.64)
Three 0 0
Four or more 0 0 Post-partum
Anemia
Baseline Hgb 120.4 (6.1) 119.5 (5.6) 0.40
* t-test, **chi-square
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
 Start with the Hypothetical Tables
Table 1. Baseline Table 2. Blood Loss and Post-partum Hgb
Carboprost Oxytocin p Value Carboprost Oxytocin p Value
Mean Age 28.13 (4.2) 28.7 (4.4) 0.42 Blood Loss 350 (116) 486 (97) <0.01
Gravidity Post-partum Hgb 119 (6) 112 (11) <0.01
One 13 (21.6%) 11 (18.3%) 0.65
Two 19 (31.6%) 16 (26.6%)
Three 28 (46.6%) 33 (55.0%)
Four or more 0 0
Parity Table 3. Prevention of PPH and PP Anemia
Zero 13 (21.6%) 11 (18.3%) 0.56 Carboprost Oxytocin RR (95% CI)
One 19 (31.6%) 16 (26.6%)
Two 28 (46.6%) 33 (55.0%) Post-partum 5 (8.3%) 24 (40.0%) 0.28 (0.12-
Three 0 0 Hemorrhage 0.64)
Four or more 0 0 Post-partum 28 (46.6%) 39 (65.0%) 0.69 (0.48-
Baseline Hgb 120.4 (6.1) 119.5 (5.6) 0.40 Anemia 0.98)
 Clinical vs. Statistical Significance
o Interpretation of the statistical results requires critical thinking, not only
on p value.
o Statistical significance can be obtained simply by increasing the sample
size, so that even a very “small effect size” can be statistically significant
but may not be clinically significant
o There is admittedly no easy way to determine clinical significance
o A rational, biological or clinical explanation for the results before
making any conclusion is necessary
Thank You
 Workshop 2
• Develop the Research Plan for Your Research
Question
• Assign a reporter from the group for presentation
Group Reporting